Hi,
Yes, of course dc coupled can sound better, but the schematic shows no way to trim the offset for each of the op amp channels in the I/V section, and the PCB seems to have coupling caps ???
It would be possible to trim the +ve input bias voltage by changing the 2k7 resistor, but this would adjust the voltage into both channels and not allow compensation for the differences between each op amp channel.
So, is the schematic the same as the board ? Any dc offset present ?
Yes, of course dc coupled can sound better, but the schematic shows no way to trim the offset for each of the op amp channels in the I/V section, and the PCB seems to have coupling caps ???
It would be possible to trim the +ve input bias voltage by changing the 2k7 resistor, but this would adjust the voltage into both channels and not allow compensation for the differences between each op amp channel.
So, is the schematic the same as the board ? Any dc offset present ?
Hi Klipschkid,
The schematic is for the Mark II version of the board provided by the seller, they might have made some changes on the the mark III PCB (the one I got). The schematic didnot show the mute section neither so I just use it for reference only.
Best regards
The schematic is for the Mark II version of the board provided by the seller, they might have made some changes on the the mark III PCB (the one I got). The schematic didnot show the mute section neither so I just use it for reference only.
Best regards
What schematic are you looking at? AFAIK there's no cap at output, fortunately.
BTW: if you do have a coupling output cap, you can measure if there's any offset before the cap. If there's less than 10mV you can bypass the cap.
Also if you have coupling caps in following stages, that output cap can be bridged (bypassed).
But muting is important in every case, as when you power the DAC on there might be an offset peak that can cause a thump on the power amp.
Carlos
Last edited:
Well, this is why I'm asking - no caps in the schematic, no way to trim offset, caps appear to be installed on the PCB, and no comments about any dc offset. I would expect DC offset would be the first thing to check if it is dc coupled - you could have hundreds of mV and damage the end equipment....
So perhaps you can clarify if your build is dc coupled,and what the offset is, or if there are dc blocking caps ?
If you want to DC couple your DAC and get rid of the coupling caps, then it's quite easy to do - have a look at this french site for what they did with the Matrix mini-i.
Tweak du DAC Matrix mini-i - Page 6 - TDG
So perhaps you can clarify if your build is dc coupled,and what the offset is, or if there are dc blocking caps ?
If you want to DC couple your DAC and get rid of the coupling caps, then it's quite easy to do - have a look at this french site for what they did with the Matrix mini-i.
Tweak du DAC Matrix mini-i - Page 6 - TDG
post #53 last pic, so far quite clear to me that it is dc coupled, but I don't have the board , I'll wait for redj and skylab confirmation.
I did made opamp board for balance-single ended conversion, you can get low dc offset by careful component matching & layout, for safety reason I did have option to use Cout though 🙂
I did made opamp board for balance-single ended conversion, you can get low dc offset by careful component matching & layout, for safety reason I did have option to use Cout though 🙂
Hi ide,
I find the alum. caps configuration odd. I don't think they can all be power decoupling. There would be 6 caps; 2 for each op amp, and all the same. It looks like the I/V has no alum local decoupling caps, the two BC136 (?) 22uf would be coupling caps, and the Elna Tonerex would be power decoupling for the buffer op amp ? I'm guessing of course.
So what offset did you get exactly ?
I find the alum. caps configuration odd. I don't think they can all be power decoupling. There would be 6 caps; 2 for each op amp, and all the same. It looks like the I/V has no alum local decoupling caps, the two BC136 (?) 22uf would be coupling caps, and the Elna Tonerex would be power decoupling for the buffer op amp ? I'm guessing of course.
So what offset did you get exactly ?
Why fight with it? If there are no caps in schematic and on photos, then it is DC coupled.
The caps that you can see on pcb look like supply decoupling, not directly in signal path.
In fact, one thing I like on this DAC is that it doesn't have caps in signal path, except for those film types on filter.
But you are certainly right on one thing: DC offset has to be checked before connecting to power amp.
Offset can damage speakers if there are no caps to block it.
Unbalancing circuit on AD datasheet specifies no output cap, but that also very much depends on the offset of the chip you are using. The chips we have been mentioning here could be used with no output caps.
those light blue philips do not sound good as coupling, tried that.Still waiting for owner confirmation 🙂. In my opamp board for tda1541 balanced I got less than 8mV using LME49720 and +/-15V rails
What light blue Philips caps are you talking about? The square small ones are film caps.
The more recent National family of opamps are mostly low-offset types, and can usually avoid caps.
My DAC will take a few weeks to get here, so I can't check on that. Can anyone check DC on output?
" the two BC136 (?) 22uf...." this one, its an electrolyte, and I personally liked oscons for decoupling
I'm using the stock board. No mods, no op amp swaps - yet. DC on output is: L=1.6mv and R=2.8mv Way to low to damage anything down stream.
I measured it at the RCA output jacks. Is that correct? But oddly enough I think my board is different than skylab's. I'm not going to remove my board at the moment so I can't check the underside, but there does appear to be 2 small yellow caps right before the jack outputs. So is it still possible to get that small of DC voltage if it is coupled through a 222mfd cap?
Hi Red,
Thanks for posting the offset info. It seems you have probably got a supply of 12.6v into the +, and that is dropped to 2.7V and fed into the - , and these two balance to create almost no offset. Nice.
Any cap will block DC so for sure they are not coupling caps - probably an hf filter.
If you can get adapters (lots on Ebay these days), then AD8620 are great for the I/V - they are designed for the +/-12V rails in the DAC and they are jfet - so they won't strain the ad1955's current output. A dip8 alternative would be OPA2134 (fet input) but they don't sound as good imo. They can actually give a deep bass boost in some applications too.
I've done a lot of op amp rolling in a PCM1793 DAC. You don't wanna get too into it ;-) There's no "perfect" op amp.
Thanks again,
Tom
Thanks for posting the offset info. It seems you have probably got a supply of 12.6v into the +, and that is dropped to 2.7V and fed into the - , and these two balance to create almost no offset. Nice.
Any cap will block DC so for sure they are not coupling caps - probably an hf filter.
If you can get adapters (lots on Ebay these days), then AD8620 are great for the I/V - they are designed for the +/-12V rails in the DAC and they are jfet - so they won't strain the ad1955's current output. A dip8 alternative would be OPA2134 (fet input) but they don't sound as good imo. They can actually give a deep bass boost in some applications too.
I've done a lot of op amp rolling in a PCM1793 DAC. You don't wanna get too into it ;-) There's no "perfect" op amp.
Thanks again,
Tom
Last edited:
Yup, that measurement is after the divider, Tom is right, there is no C out, only filter
Thanks for checking it
Thanks for checking it
So we are talking supply decoupling then?
There are many parts on this DAC that might be improved, and that's one of the things to do if you feel like it.
Unfortunately the sellers only provide assembled DACs, so a little unsoldering would be in order to swap parts.
yes, that's unfortunate to unsolder things. but I have did worst pcb than this, the one with SMD's
For now I was limiting my potential mods to things you could unplug or add to what it's already there.
Like the socketed ICs, an external raw supply or an input digital transformer.
By doing so there might be more people willing to experiment things. Unsoldering/soldering parts is not something everyone wants to do. 😉
Like the socketed ICs, an external raw supply or an input digital transformer.
By doing so there might be more people willing to experiment things. Unsoldering/soldering parts is not something everyone wants to do. 😉
- Status
- Not open for further replies.